Clinical characteristics of familial generalized anxiety disorder

The authors seek to determine whether the clinical characteristics of generalized anxiety disorder (GAD) differ in individuals with a high vs. low familial vulnerability to illness. We identified 486 personally interviewed female twins from a population‐based register who had both an interviewed co‐twin and a lifetime history of GAD using modified DSM‐III‐R criteria which required a one‐month minimum duration of illness. We attempted to predict risk for GAD in the co‐twin from the clinical features of the GAD in the proband twin using the Cox proportional hazard model, controlling for year of birth and zygosity. Only two variables uniquely predicted an increased risk for GAD in the co‐twin: number of GAD symptoms endorsed and comorbidity with bulimia. Variables that did not uniquely predict risk of illness in the co‐twin included age at onset, duration of the longest episode and number of episodes. The familial vulnerability to GAD can be meaningfully indexed by clinical features of the syndrome. These results suggest that if the syndrome of GAD is to be narrowed, it would, from a familial perspective, be more valid to increase the minimum number of required symptoms rather than to increase the minimum duration of illness. Anxiety 1:186–191 (1994/1995). © 1995 Wiley‐Liss, Inc.